Trimethine Cyanine Dyes as NA-Sensitive Probes for Visualization of Cell Compartments in Fluorescence Microscopy

Abstract

We propose symmetrical cationic trimethine cyanine dyes with β-substituents in the polymethine chain based on modified benzothiazole and benzoxazole heterocycles as probes for the detection and visualization of live and fixed cells by fluorescence microscopy. The spectral-luminescent properties of trimethine cyanines have been characterized for free dyes and in the presence of nucleic acids (NA) and globular proteins. The studied cyanines are low to moderate fluorescent when free, but in the presence of NA, they show an increase in emission intensity up to 111 times; the most pronounced emission increase was observed for the dyes T-2 in the presence of dsDNA and T-1 with RNA. Spectral methods showed the binding of all dyes to nucleic acids, and different interaction mechanisms have been proposed. The ability to visualize cell components of the studied dyes has been evaluated using different human cell lines (MCF-7, A2780, HeLa, and Hs27). We have shown that all dyes are cell-permeant staining nucleus components, probably RNA-rich nucleoli with background fluorescence in the cytoplasm, except for the dye T-5. The dye T-5 selectively stains some structures in the cytoplasm of MCF-7 and A2780 cells associated with mitochondria or lysosomes. This effect has also been confirmed for the normal type of cell line-human foreskin fibroblasts (Hs27). The costaining of dye T-5 with MitoTracker CMXRos Red demonstrates specificity to mitochondria at a concentration of 0.1 μM. Colocalization analysis has shown signals overlapping of dye T-5 and MitoTracker CMXRos Red (Pearson's Coefficient value = 0.92 ± 0.04). The photostability study shows benzoxazole dyes to be up to ∼7 times more photostable than benzothiazole ones. Moreover, studied benzoxazoles are less cytotoxic at working concentrations than benzothiazoles (67% of cell viability for T-4, T-5 compared to 12% for T-1, and ∼30% for T-2, T-3 after 24 h). Therefore, the benzoxazole T-4 dye is proposed for nucleic acid detection in vitro and intracellular fluorescence imaging of live and fixed cells. In contrast, the benzoxazole dye T-5 is proposed as a good alternative to commercial dyes for mitochondria staining in the green-yellow region of the spectrum.

Figure 1

Chemical structure of studied symmetrical trimethine cyanine dyes T-1–T-5.

Figure 2

Absorption spectra of the trimethine cyanine dye T-1 in dsDNA (a) and RNA (b) presence in 0.05 M Tris-HCl (pH 7.9). Dye concentration 5 μM.

Figure 3

Fluorescence imaging of live (a) and fixed (b) MCF-7 cells stained with the studied dyes T-1–T-5 (dyes concentrations for live cell imaging: T-1–10 μM; T-2, T-3–1 μM; T-4, T-5–0.1 μM; for fixed cell imaging: T-1, T-4–1 μM; T-2, T-3, T-5–0.1 μM). Scale bars: 10 μm.

Figure 4

Comparisons of the cytotoxicity of trimethine cyanines T-1–T-5 at different concentrations (0.1, 1, and 10 μM) toward A2780 cells for 6 (a) and 24 (b) h. Working concentrations are marked with a circle. The standard deviation is presented as error bars.

Figure 5

Fluorescence imaging of live Hs27 cells stained with the studied dyes T-1–T-5. (T-1 at the concentration of 10 μM, T-2, and T-3 dyes at the concentration of 1 μM, T-4, and T-5 at the concentration of 0.1 μM). Scale bars: 10 μm.

Figure 6

Time-dependent confocal fluorescence microscopy imaging of HeLa (a) and Hs27 (b) cells stained by trimethine cyanine dyes T-1–T-5 (at the concentration of 10 μM for T-1, 1 μM for T-2, T-3, 0.1 μM for T-4 and T-5 for Hs27, and the same for HeLa, but the concentration for T-3 is 0.1 μM) under constant irradiation for 8 min.

Figure 7

Fluorescence live-cell imaging of Hs27 cells. Colocalization analysis of the dye T-5 (green, a) with MitoTracker CMXRos Red (red, b) and merged image (c). Scale bars: 10 μm.